The Laboratory has undertaken a reinvestigation of the use of bacterial viruses (phage) for the treatment of bacterial infections. The discovery of viruses which can infect and destroy bacteria was greeted with considerable optimism at the beginning of this century. One of the discoverers of these viruses (bacteriophage) Felix d'Herrelle traveled throughout the world promoting their use as therapeutic agents for the treatment of infectious diseases. Despite his enthusiastic efforts, which were widely known at that time the use of these bacteriophage as therapeutic agents has in general been abandoned by the medical community. In part the failure of bacteriophage to gain general acceptance as a therapeutic agent was due to the greater efficacy displayed by the antibiotics following their introduction in the 1940s. In reviewing the literature concerning attempts to use phage for the treatment of bacterial infections it became clear that there were three fundamental problems which may have hindered progress. First, the pioneers in this field thought that a single strain of phage could be used to treat numerous strains of bacteria. We now know that the phage are generally highly specific. Second, there was very limited recognition concerning the problems generated by the bacterial debris (endotoxins) which accompany phage production. We recognized and addressed this problem by purifying the phage with cesium chloride density centrifugation. Third, research previously performed in this Laboratory demonstrated that normal or wild-type phage are often rapidly removed from the circulation by the reticulo-endothelial system. To overcome this problem we developed a method for the selection of phage which are not removed as rapidly from the circulation. In preliminary experiments with phage preparations, which meet the 3 criteria noted above, we have been able to rescue animals injected intraperitoneally with lethal doses bacteria. In addition, we have shown by high resolution two-dimensional protein electrophoresis that the phage selected to avoid capture by the reticulo-endothelial system differ from the parent wild-type phage by a mutation in the major coat protein of the phage.